Data::Domain

1
DataDomain

• a data validation tool
• laurent.dami_at_justice.ge.ch

2
What is a "data domain" ?

• term from data management
• a set of values
• may be infinite
• defined by extension (enumeration) or by
  intension (set of rules)

3
How to work with domains ?

• To put the definition into practice, we need to
  be able to
• Define a domain
• atomic building blocks (mostly scalar)
• composition operators
• Check if a value belongs to a domain
• if not explain WHY
• answers should be consistent over time
• validating a withdrawal from an account is
  not a domain operation

4
Why data domains ?

• when some data crosses "boundaries"
• user form
• database
• parse tree
• config. file
• function call
• principle of defensive programming

5
CPAN many modules

• Parameter checking
• ParamsCheck, ParamsValidate
• Data Modelling Object-Relational Maps
• JiftyDBI, Alzabo, RoseDBObject,
  DBIxClass
• HTML Form tools
• CGIFormBuilder, DataFormValidator
• Business rules
• Brick, DeclareConstraintsSimple,
  DataConstraint
• Terminology "domain" often called "template" or
  "profile"

6
Other technologies

• database validation mechanisms
• reference table
• rules constraints
• triggers
• typing (strong / dynamic)
• XML schema
• Javascript frameworks
• Parsers
• ...

7
Some design dimensions

• shape of data
• scalar (string, num, date, ...)
• array or hash
• multi-level tree
• objects
• shape of messages
• single scalar
• collection
• multi-level tree
• Conciseness (declarative style)
• Expressiveness
• Internal dependencies (i.e. begin_date /
  end_date)

8
Scenario
Ajax submit
Key Value
Foo.1.bar
Foo.2.buz
CGIExpand
table
HTML form
data tree
DataDomaininspect
decorate
invalid
JSON error messages
valid
process form (?DBIxDataModel)
see video
9
Synopsis

• my domain Struct(
• anInt gt Int (-min gt 3, -max gt 18),
• aNum gt Num (-min gt 3.33, -max gt 18.5),
• aDate gt Date(-max gt 'today'),
• aLaterDate gt sub
• my context shift
• Date(-min gt context-gtflataDate)
• ,
• aString gt String(-min_length gt 2,
• -optional gt 1),
• anEnum gt Enum(qw/foo bar buz/),
• anIntList gt List(-min_size gt 1, -all gt
  Int),
• aMixedList gt List(Integer, String, Date),
• )
• my messages domain-gtinspect(some_data)
• display_error(messages) if messages

10
Design principles

• Do One Thing Well just check
• no HTML form generation
• no Database schema generation
• no data modification ( filtering, canonic form)
• return informative messages
• concise yet expressive
• extensible (OO inheritance)

11
Domain creation

• Object-oriented
• my dom DataDomainString-gtnew(
• -min gt "aaa",
• -max_length gt 8,
• -regex gt qr/foobar/,
• )
• Functional shortcuts
• my dom String(-min gt "aaa", ...)
• Default argument for each domain constructor
• my dom String(qr/foobar/) default is
  -regex
• Arguments add up constraints as "and"

12
Generic arguments

• -optional
• if true, an undef value is accepted
• - name
• name to be returned in error messages
• - messages
• ad hoc error messages for that domain

13
Builtin scalar domains

• Whatever (-defined, -true, -isa, -can)
• Num, Int (-min, -max, -range, -not_in)
• Date, Time(-min, -max, -range)
• String (-regex, -antiregex, -min, -max,
  -range, -min_length, -max_length,
  -not_in)
• Enum (-values)

14
Builtin structured domains

• List (-items, -min_size, -max_size,
  -all, -any)
• Struct (-fields, -exclude)
• One_of (-options)

15
Example

• use RegexpCommon
• sub Name
• return String(-regex gt qr/-.
  alpha/,
• -antiregex gt qr/REprofanity/,
  
• _at__)
• my person_dom Struct(
• lastname gt Name,
• firstname gt Name(-optional gt 1),
• d_birth gt Date(-optional gt 1,
• -max gt 'today'),
• )

16
Lazy Domains

• Principle
• a coderef that returns a domain at the time it
  inspects a value
• can look at the surrounding context (subvalues
  seen so far)
• my person_dom Struct(
• ...
• d_birth gt Date(-optional gt 1,
• -max gt 'today'),
• d_death gt sub
• my context shift
• return Date(-min gt context-gtflatd_birth)
  
• ,
• )
• 
  (inspiration ParseRecDescent)

17
What is in the "context"

• root
• top of tree
• path
• sequence of keys or array indices to the current
  node
• list
• ref to last array visited while walking the tree
• flat
• flattened hash with all keys seen so far

18
Example Contextual sets
my some_cities Switzerland gt qw/Genève
Lausanne Bern Zurich Bellinzona/, France
gt qw/Paris Lyon Marseille Lille Strasbourg/,
Italy gt qw/Milano Genova Livorno Roma
Venezia/, my domain Struct( country gt
Enum(keys some_cities), city gt sub
my context shift my country
context-gtflatcountry return
Enum(-values gt some_cities-gtcountry) ,
)
19
Example Ordered list
my domain List(-all gt sub my context
shift my index context-gtpath-1
return Int if index 0 first item my min
context-gtlistindex-1 1 return
Int(-min gt min) )
20
New Domain Constructors

• by wrapping
• sub Phone
• String(-regex gt qr/\?0-9() /,
  -messages gt "Invalid phone number",
• _at__)
• by subclassing
• implement new()
• implement _inspect()